Face Liveness Detection Using CNN for Face Verification on Medical Robot

Abstract

The development of science in the field of technology has accelerated due to the global need for society 5.0 which emphasizes creativity, critical thinking, communication, and collaboration to solve existing problems. One of its applications is to collaborate to create a smart robot using computer vision that can recognize a person's face and verify its authenticity. face recognition uses the identification and classification of faces by comparing matrix values between models and new inputs. Face liveness detection on the robot uses the mediapipe and the CNN algorithm. Facial recognition results get an accuracy value of 92% with an optimal distance of 100 cm.